Wednesday March 4th 2009. Launch -43 days. 3 days to golf match with NASA.

 

A lot has been happening. First of all a lot of quiet checking has been going on about the launch status of Herschel. Pre-flight check-out is going extraordinarily well. At present we are 5 days ahead of schedule and looking in great shape. Unless something unexpected happens, Herschel will be ready to launch on time. That is not to say that there is no possibility of a delay: a launch is so complex that it can be delayed for any one of thousands of different reasons – it is hardly something for the nervous to take part in a launch campaign.

 

While the specialists in Kourou are getting the Ariane 5 launcher, Herschel and Plank ready for launch a lot of work is going on getting systems ready to handle the data when it comes down. Astronomers will only be aware of the huge network that gets their data to them from the satellite if it should go wrong somewhere. Our job is to try to ensure that it is as robust as possible and does not go seriously wrong. Over the last few weeks the main concern has been to get the launch version of the software ready. This is a huge and exceptionally complex package that includes:

 

 

Some of this software has been used and thoroughly tested for a long time. A lot of it, particularly for data distribution and data processing, is new. It all goes through stage after stage of rigorous testing but, inevitably, the only real test is to throw it to the wolves – that is, real astronomers, who will always find ways to break it that we have never dreamed of. The reason is quite obvious: it is one thing to use software in a carefully controlled way following standard procedures, but astronomers will try it out in imaginative and unexpected ways with all kinds of data. This means that we have to make the software extremely robust, quite apart from finding and fixing the hundreds of bugs that any sophisticated piece of software will have, however carefully it is written. Bugs come in all shapes and sizes from the little irritations that can be got around to the worst kind, known as “showstoppers” because, unless fixed, the show will not go on. A major software release starts by preparing an initial version that seems to work; this is a release candidate. The release candidate is checked out in system testing and any show-stoppers are identified and studied. Until showstoppers are found and fixed the software is not deemed suitable even for further testing. Only when the showstoppers are fixed and the system testing is successful does a release candidate become a release and goes through extensive testing. Over the last three weeks we have gone through a series of release candidates. Finally, number four has turned out to pass all the initial checks and will now be worked on intensively. From here we will look for second-level bugs and fix them, issuing patches until we have a version of the software that we are happy with. This will then be frozen for launch.

 

The Videocon with the Instrument Specialists this afternoon confirmed that after a very nervous two weeks spectacular progress has been made in the last 7 days. Whereas last week we were seriously concerned that we might not have the software in a suitable state to use for the Simulations Campaign next week, now everyone is convinced that it is ready to go. The Videocon was positive and almost all good news.

 

Less successful was our attempt to get some vital practice in for our golf match v NASA on Saturday. Four of our team turned up at the course this afternoon to be greeted by a strong gale and hail. There was fresh snow on the hills behind. After spending some 20 minutes warming up – if you can call it that when the temperature is 6 degrees and dropping fast and there is driving rain and hail – a new and very intense shower sent us scurrying for cover. After a few minutes debate, we decided that even we were not mad enough to try and practice in such awful conditions and gave up. We will try again on Friday evening.

 

Although it was not raining at work when I rode home finally, there was a strong wind. In the morning I was riding up hill into a stiff breeze. This evening I had a gale either in my face or as a crosswind. Several times it became difficult to stay upright. The long climb up the last few kilometres became a painful crawl in low gear and it was a relief to do it in daylight: in the dark it would have been nerve-wracking. Right now it is late and I can hear heavy rain outside. The ride to work in the morning may be fun…